384 ALASKAN GLACIER STUDIES
in about the same position from 1899 to 1909, as indicated by comparison of the photographs by the Harriman Expedition and by Grant and Higgins. At the western margin the advancing ice was overriding shrubs of willow and alder, and there were ice blocks sliding from the lateral cliffs into the alders; but higher up on the mountain slope the western margin of the glacier had not yet overridden the whole of its narrow barren zone; and there was still a barren zone on the eastern side extending from 100 to 700 feet beyond the glacier. At the outer edge of this barren zone is a small lateral moraine.
Harriman GlacieróGeneral Description. Harriman Glacier, which is at least 9 miles long and about a mile wide at the terminus, flows northeastward into the extreme southern end of TTfl.rriTnfl.n Fiord. Its whole surface is moderately crevassed, and it is quite free from moraines, there being two medial moraines near the southern side but no lateral moraines. The terminus is a vertical cliff 200 to 800 feet high, from which many icebergs are discharged. Six or seven hundred feet from the middle of the ice front the water is only 90 feet deep.
N$D&-Sheathed Slopes. Harriman Glacier differs from most of the others in the Port Wells region in the nature of its tributaries. Instead of being fed by a number of well-defined tributaries, occupying subordinate valleys and each with a length of several miles, it is fed by a considerable number of rather short cascading tributaries, from cirques in peaks 5815 to 8206 feet high. The longest tributary seen from Harriman Fiord is not over a mile or two in length. A great deal of the ice is supplied from neve1 fields between these short tributaries. These neve1 fields (PI. CXXXVI) are intermediate between the snowfields arid the ice tongue, and around the borders of Harriman Glacier, large areas of mountain slope are completely sheathed with such neve ice. On the sketch maps by Gannett and by Grant and Higgins these areas of n6v£ are not specifically distinguished from true ice tongues; and it would indeed be impossible to map them accurately without the most detailed work in the nev6 area.
The phenomenon of ice-sheathed or neve-sheathed mountain slopes, developed on a large scale in Harriman Fiord, is a common feature of Alaskan glaciation of the present time found in other areas, as in Passage Canal, along the eastern side of Lynn Canal in southeastern Alaska, and doubtless in many other localities where the snowfall is great and the snow line is rather low. If the snowfall should decrease slightly it seems likely that these neVe-sheathed slopes would become bare and that the short tributaries would shrink and become better-defined, though smaller, ice tongues than at present.
Retreat. The front of Harriman Glacier has oscillated considerably since 1899 when it was described by Gilbert and mapped by Gannett. Previous to that year there seems to have been some retreat, for Gilbert observed that the glacier was not "closely approached by forest growth, but shrubs were seen on the shore of the fiord within a few hundred yards of the ice." This suggests a narrow barren zone.
Grant and Higgins state that "photographs of the eastern side of the front of Harriman Glacier taken from Point H (Fig. 50) in 1905 and 1909 show that this side of the glacier (eastern) retreated approximately 700 feet between these dates. A comparison of an 1899 photograph with the above indicates that between 1899 and 1905 the east side of the glacier retreated about half the above distance. As the two photographs were not taken from the same point this estimate of the retreat between 1899 and 1905 is only approximate. On the west side of the ice front a careful examination in 1909 of the glacier from the position of a photograph taken in 1899 showed no noticeable difference